Modeling effective factors on aggressive behavior of taxi drivers using Logistic Regression – Case Study: Tehran Metropolitan

Aggressive behavior of drivers is one of the most important issues in the behavioral and social sciences. The importance of this issue is due to its direct relationship with the violations of drivers and the safety of other users. In addition, taxi drivers spend too much hours during a day in the urban streets. This makes them more susceptible for anger and aggressive driving. Analyzing the aggressive behavior of drivers requires identifying the factors and conditions that affect them. So far, the way to collect data to determine the effective factors on aggressive behavior of drivers has been through designed questionnaires, which have been completed by the drivers themselves. Nevertheless, this method itself has limitations on the reliability and accuracy of the answers given. Therefore, in this paper it has been attempted to obtain data with the help of a mobile-based application, which would be completed by an informed passenger and without letting driver to know the procedure. Due to the nature of the dependent variables of this study which their scale is zero and one (occurrence or non-occurrence of any of the dependent variables); prediction of factors affecting aggressive behavior of taxi drivers has been performed using multiple logistic regression method. Among the most important factors identified to provoke anger in taxi drivers are slow-moving of leading vehicles with respect to traffic flow with effective coefficient of 2.95, sudden redirection or lane changes performed by other vehicles with effective coefficient of 2.817, sudden, slow or illegal pedestrian crossing across the street with effective coefficient of 2.717, failure in observance of proper safety interval between vehicles with effective coefficient of 2.488, lack of feeling safety during travel with effective coefficient of 2.394 and the occurrence of accident and it’s probability along the route with effective coefficient of 1.932.